An efficient approach for Interactive Sequential Pattern Recognition

作者：

Highlights：

• Searching the optimal decoding for Interactive Sequential Pattern Recognition tasks may be computationally expensive.

• A normalization procedure is proposed so that computing optimal hypotheses becomes more efficient.

• A specially profitable case is reported when dealing with Regular Languages.

• Experimental results show a significant save of computational time.

摘要

Highlights•Searching the optimal decoding for Interactive Sequential Pattern Recognition tasks may be computationally expensive.•A normalization procedure is proposed so that computing optimal hypotheses becomes more efficient.•A specially profitable case is reported when dealing with Regular Languages.•Experimental results show a significant save of computational time.

论文关键词：Interactive Pattern Recognition,Sequential Pattern Recognition,Human-computer interaction,Word graphs,Efficient search

论文评审过程：Received 27 June 2015, Revised 7 November 2016, Accepted 8 November 2016, Available online 14 November 2016, Version of Record 2 December 2016.

论文官网地址：https://doi.org/10.1016/j.patcog.2016.11.006